[gwood]

See Attached Files for A2100 Batteries and Power Supply info...

[Kwirk]

I had a change to play abit more.
I have a part where i circular milled diameter 296mm and diametee turns out 0.04mm smaller in Y direction (0.02mm per side)

I checked the balance and gain,my machine is in swedish so not sure if i got it right. I adjusted "balanskompensering" til KV was same in both directions

I did notice the X had value +0.72 (adjusted to 0.6)
And Y axis -1.12. (Adjusted to -0.65)

This made no difference in the circularity

I wonder if this has to so with my squareness error between X and Y but cant figure out what happens really?

I cant feel any mechanical backlash where machine change direction.

If i mill a small hole it is good.

Does the 0.013mm squareness error across 500mm cause this?

[Kwirk]

Ok i might have measured abit sloppy.

I would say a diameter of 300mm is 0.02mm smaller in Y direction.

Meaby it was abit better after the Kv got dialed.
I have found something i think
In the pitch error compensation Y axis has really small values. Like 0.002

The X axis has bigger values across the table. Something like 0.012

Could it be that it is over compensated in X?
Why would one put values in there when it has scales?
Meaby one time the machine was sloppy in X axis..
I have replaced the coupling at the motor and motor has been overhauled by siemens tech sometime.

I dont really wanna mess with the table as i dont understand how it works, gwood do you know?
I see it has increasing values from 0 to 510 and then values to put + - and so on...

[gwood]

Kwirk,
See Attached file for Table Flatness and Tram. This needs to be checked First.
Make your tests with a cold spindle at start up.
The Tables were done in the Factory with a level machine table using a laser measuring device.
A ball-screw or linear scale may not be exactly parallel to the linear ways producing errors.
The Y-Axis also has Thermal Compensation due to the fact that the Spindle Carrier grows out and away from the carrier as it heats up.

[Kwirk]

Thanks
I would say the tram is good.(it was atleast lets see)
When i got the machine it was really bad in X direction
So i spent alot time dialing it in. I did this at fairly big diameter. 350/380mm if i remember right..i think i got it within 0.02 difference. Good enough?

I must say it was sensitive in the tram in Y direction with leveling feets compared to my old B.P.vmc.

But i will definately check the tram again as it has sa on the floor couple of months now.
I did some face milling with 50mm mill and the "waves" between then passes were really "good" . Minimal.

I have also done a test where i find center of a ring with shortest possible holder with dial mounted in

Then change to longest possible..like 350mm longer
And the center is within 0.015 0.02 .

[gwood]

normally I prefer .012mm / 300mm (.0005" / 12" inches). If it is out more in X, than Y this would cause what you see.
Also with scales you cannot check backlash since the scale is trying to hold the axis in position.
This is NOT the case under a dynamic load when cutting!
You need to look at the ball-screws. (Procedure attached)
"_Troubleshooting_Axis_Positioning_Problems-VMC-110518GEW.pdf" (289.0 KB)

[Kwirk]

normally I prefer .012mm / 300mm (.0005" / 12" inches). If it is out more in X, than Y this would cause what you see.
Also with scales you cannot check backlash since the scale is trying to hold the axis in position.
This is NOT the case under a dynamic load when cutting!
You need to look at the ball-screws. (Procedure attached)
"_Troubleshooting_Axis_Positioning_Problems-VMC-110518GEW.pdf" (289.0 KB)

Thanks alot for taking your time

I havent checked the screws more then by feeling them byhand
And having a 0.002 dial gauge. They feel really nice.
The finishes it leaves during roughing and finishing are really nice. I cant even feel the slightest errors where it changes direction so focus on the screws has been minimal.

I have witnessed some sloppy screws on the scales machines and during machining you can easily feel the backlash on circles

I am thinking of stacking up some calibration gages of 300mm measure first in X with 3d taster. Then measure in Y direction ( first turn taster 90 degreees )
Im pretty sure the measurement will differ 0.02.

What would be the cause of this then?

[gwood]

The tests I provide instructions for on the end of the ball-screw take less than 15 minutes;
spend your time looking for bad thrust bearings; or ball-screws that have lost their preloads.

[Kwirk]

Work getting in the way of perfecting this old girl.
I did have a feel on the backlash i would say its in the 0.005mm so imo. Still pretty tight.

I have confirmed that a dia. 44mm hole is interpolated within 0.005mm

Dia 170. In 0.015mm

Dia 300 0.035mm

I did end up squaring the X and Y from 0.02mm to 0.008 it gave
Very little error.

I have confirmed against a square granite that the table dont drop towards X limits.

I have milled 4 flats 300x300 and the difference in measures across is just the same as the circle..Smaller in Y

I have done some testing how it repeats positioning and its very very good. Nothing seems to be loose.

I have tried turning off pitch error compensation in X and Y and it gave nothing.

My next step is to mount a separate 0.001mm resolution scale on the table with holder from spindle and compare the movements.

Unless anyone comes up with something more to check

[Kwirk]

Another thought
Does anyone know where to access the thermal compensation for Y? The machine is in a enviroment that is abit colder then 20c wondering if that could mess things up?
But i reckon it should compensate for "float" of spindle heat on the column rather then stroke?

[gwood]

Kwirk, There are two thermocouples; one on the Side of the column near the toolchanger assy., the other inside the spindle carrier down on the inside where the Spindle assembly bolts in.
The Spindle assy has a 13 digit 25mm high plate riveted on the bolting flange outer diameter 06126xxxAyyyy below the casting; but above the lower sheet metal cover.
The MSN 70xx-A0x-yyyy
Does the machine have a spindle heat exchanger and fluid pump.
What is the MAX programmable spindle speed?
On the A2100 Service page what release software is loaded? System Software = SYyymmdd Release 2.x, 3.x, 4.x ?
This info is need to verify the Thermal comp tables.

[Kwirk]

Thanks for getting back

The spindle nr is 061264424000
It is a 8k unit with no chiller. Oil lubricated.

Software release 4.13
MV001214

[gwood]

Kwirk,
Attached: 91204316-001A_EKK_Rel4x_Config.pdf (354.9 KB) Page # 2 & # 5 (Thermal Comp)
Your Spindle (61264424A 8k STD Torque) or 10K with Spindle Cooler
Attached: VMB-TB-KT-012a_EKK_ILC_Lube_Setup-Rev_A.pdf (386.3 KB)
GWood

[Kwirk]

Kwirk,
Attached: 91204316-001A_EKK_Rel4x_Config.pdf (354.9 KB) Page # 2 & # 5 (Thermal Comp)
Your Spindle (61264424A 8k STD Torque) or 10K with Spindle Cooler
Attached: VMB-TB-KT-012a_EKK_ILC_Lube_Setup-Rev_A.pdf (386.3 KB)
GWood

Thanks. Cant see anything wrong

Im thinking off turning the "thermal comp selection" feom ambient to inactive or absolute. Should be safe ?

And then do a test.
Control re boot needed?

[gwood]

It your choice; but I would bet it is mechanical alignments, or leveling.

[Kwirk]

It your choice; but I would bet it is mechanical alignments, or leveling.

I believe you..just running out of ideas.
I did now measure with 0.0001" dial and the out of round is
0.0127 per side so a diameter in X is 284.00m and in Y
283.9873mm

The smaller it get the better it is.

Would a tram error do this ?
Would tram error result a square to be smaller in one direction too ?

I will borrow a level and check the machine when i have time for it

[Kwirk]

For now i adjusted pitch error til the Y axis makes the perfect circle and same size square as X
It now interpolates a d250mm within 0.005mm.
Makes square 250x250
A smaller hole is in same center location as 250.

When cmm time at work is avaible i can fine adjust other stuff.

I feel like i did a dirty fix as its probably something
Mechanical as gwood states.